Committee

Prof. Naresh Thadhani – School of Materials Science and Engineering (advisor)

Prof. Min Zhou– School of Materials Science and Engineering

Prof. Blair Brettmann – School of Materials Science and Engineering

      

Prof. Arun Gokhale – School of Materials Science and Engineering

Prof. Brian Jensen – School of Physics and Astronomy, Washington State University

Abstract

High-solids loaded polymer composites contain several hierarchies of heterogeneities and are of interest for use as ceramic green bodies and energetic crystals embedded in a polymer matrix. The recent and rapid growth of additive manufacturing (AM) and the engineering need for more complex geometries and individualized products has led to a surge of interest in fabricating high-loading particle composites via AM. In particular, Direct Ink Write (DIW) extrusion involving layer-by-layer deposition of a composite paste made of a highloading of solids and a curable polymer binder is used to fabricate such composites in different geometries and forms. The layers can be deposited to form various structures, such as the colinear structure (filaments and layers are all parallel) and the log-cabin structure (filaments in a single layer are all parallel, each layer is perpendicular to the previous one). This technique is very convenient for fabricating composite structures, since the highly viscous nature of the high-solids loaded paste can be countered with the use of shear-thinning binders and there are no heat effects to consider. However, DIW-AM introduces further complexity in composites due to formation of process-inherent heterogeneities such as particle aggregation or porosities, which can be random, directional, or stochastic. The structure and composition of such materials vary across several length scales, resulting in processing and mechanical behavior that is difficult to predict or understand. Shock-compression of heterogeneous particle-filled polymer composites often involves complex interactions, which can make it difficult to predict their dynamic mechanical properties. The shock compression behavior is often dominated by mesoscale defects (including porosity) or interactions of the shock wave with interfaces and particulates. xvi Traditional diagnostic methods, such as velocity interferometry, enable temporally resolved measurements, but are limited in spatial resolution and generally provide volumeaveraged responses. Spatially resolved measurements are therefore also necessary to provide sufficient information regarding the mesoscale processes which dominate performance of such materials. X-ray phase contrast imaging, a spatially and temporally resolved technique, in conjunction with traditional velocimetry, can enable observation of the effects of hierarchical heterogeneities on shock compression response. In this work, the effect of print geometry and porosity (process inherent heterogeneities) on the shock compression response of an additively manufactured high-solids loaded composite is studied. The composite contains three reinforcing phases: two inorganic particles and one organic particle, all with differing size distributions and morphologies. They are surrounded by a UV-curable polymer binder. In order to investigate the effect of these process inherent heterogeneities on shock response, the high-solids loaded composite’s microstructure is first quantitatively characterized via microcomputed tomography imaging and computational analysis in three dimensions. Next, the composite undergoes plate-impact experiments at Argonne National Laboratory’s Advanced Photon Source’s Dynamic Compression Sector, with X-ray PCI used as an in-situ and in-material diagnostic. This is combined with PDV for validation. The phase contrast images are analyzed in order to measure shock and particle velocities directly from the translation of the shock wave and particles over time. Finally, the effects of print geometry, impact direction relative to print orientation, and porosity are studied by combining the aforementioned structural characterization with the shock response of the material determined via X-ray PCI. This reveals that print geometry xvii does result in differing macroscale shock response (quantified with EOS), and that print geometry, impact orientation, and pore morphology all have an effect on microscale shock response (quantified with pore collapse velocity). We expect that these factors, only studied on a relatively small scale in this work, will become more exaggerated as sample size and therefore quantity of heterogeneities grows.